A Systematic Analysis of Metal and Metalloid Concentrations in Eight Zebrafish Recirculating Water Systems

Abstract

Metals and metalloids are integral to biological processes and play key roles in physiology and metabolism. Nonetheless, overexposure to some metals or lack of others can lead to serious health consequences. In this study, eight zebrafish facilities collaborated to generate a multielement analysis of their centralized recirculating water systems. We report a first set of average concentrations for 46 elements detected in zebrafish facilities. Our results help to establish an initial baseline for trouble-shooting purposes, and in general for safe ranges of metal concentrations in recirculating water systems, supporting reproducible scientific research outcomes with zebrafish.

Introduction

The earth's crust is composed of ∼92 elements,¹ 16 of which are essential for living organisms, 9 of which are metals or metalloids: boron (B), calcium (Ca), copper (Cu), iron (Fe), magnesium (Mg), manganese (Mn), potassium (K), sodium (Na), and zinc (Zn).² With respect to nutrition, availability and required concentrations for physiological and biological processes are used to distinguish and define these elements as macronutrients (Ca, Mg, Cl, K, and Na) or micronutrients (cobalt [Co], Cu, Fe, iodine [I], Mn, molybdenum [Mo], selenium [Se], and Zn).^3,4 The remaining elements have either no known biological function or are classified as nonessential because of their well-known toxicity or lack of evidence for roles in biological functions.^2,5

Metal uptake in fish is more diversified than in other species, and various elements are absorbed and passed into the bloodstream through gills (waterborne),^4,6,7 intestines (foodborne),^4,7 and skin (passive diffusion).^4,8,9 Furthermore, specific cell-membrane transport mechanisms transfer these elements into cells to make them available as essential components for biological processes or simply to accumulate them in specific organs such as liver, gills, or kidney.^10–12 Portions of these metals are excreted into the water, mainly through the gills and via the liver (bile) and kidney (urine).⁴

Biological problems arise when the concentrations of micro- and macronutrients are insufficient or exceed their required minimal or maximal physiological range. Elemental deficiencies usually lead to the slowing or the breakdown of physiological processes, causing various disease symptoms.^13,14 Conversely, high concentrations can also lead to severe symptoms and disorders in reproduction, growth and development,^13,15–17 metabolism,^10,18,19 and behavior.^20–22 The toxicity of particular metals is typically defined by their concentration of free ions in water, which is highly dependent on environmental parameters such as water hardness, conductivity, temperature,^23–26 and pH.^27,28

Environmental agencies such as the Environmental Protection Agency (EPA, USA) and the Australia, New Zealand Environment, and Conservation Council (ANZECC) developed guidelines for freshwater and saltwater levels for various metals (Table 1). However, innumerable organic and inorganic pollutants still lack reported toxicity thresholds for different aquatic organisms. In addition, many countries have not established their own water criteria guidelines for aquatic life²⁹ and rely on EPA or ANZECC standards.

Table 1.

Aquatic Water Criteria for Freshwater Species from the Environmental Protection Agency (USA) and the Australia and New Zealand Environment and Conservation Council

Elements	EPA		Australia/New Zealand
Elements	Freshwater CMC (μg L⁻¹)	Freshwater CCC (μg L⁻¹)	Concentration/level of protection (80%–99% species) (μg L⁻¹)
Aluminum (pH >6.5)	—	—	27–150
Arsenic	340	150	1–360 (AsIII); 0.8–140 (AsV)
Barium	—	—	—
Beryllium	—	—	—
Boron	—	—	90–1300
Cadmium	1.80	0.72	0.06–0.8
Cesium	—	—	—
Chromium	570 (CrIII); 16 (CrVI)	74 (CrIII); 11 (CrVI)	0.01–40 (CrVI)
Cobalt	—	—	—
Copper	—	—	1–2.5
Gallium	—	—	—
Iron	—	1000	—
Lead	82	3.20	1–9.4
Manganese	—	—	1200–3600
Mercury	1.4	0.77	0.06–5.4
Molybdenum	—	—	—
Nickel	470	52	8–17
Rubidium	—	—	—
Selenium	—	—	5–34
Silver	3.2		—
Strontium	—	—	—
Thallium	—	—	—
Thorium	—	—	—
Vanadium	—	—	—
Zinc	120	120	2.4–31

Maximal water concentrations for metals and metalloids according to the aquatic water criteria for freshwater species of EPA and ANZECC agencies. Hyphen: lack of information. The trigger values from EPA and ANZECC come from multiple-species toxicity tests and are not specific to zebrafish. Sources: Refs.^76,77

ANZECC, Australia and New Zealand Environment and Conservation Council; CCC, Criterion Continuous Concentration; CMC, Criterion Maximum Concentration; EPA, Environmental Protection Agency.

Zebrafish (Danio rerio), a well-established model organism in biomedical sciences, is an example for which there is a significant lack of information about acceptable concentrations and ranges of metals for animal well-being, health, husbandry, and breeding performance.^{13,15,30–34} Centralized recirculating water systems (CRWS) are widely used for zebrafish husbandry and colony maintenance, because they offer control of water chemistry parameters and concentrations of elements in the water. CRWS typically have three key components: the water source, the recirculating water filtration system (RWFS), and the aquatic housing system.

There are typically three main sources of water: groundwater, surface water, and municipal water. These sources can differ seasonally in their quality by carrying varying levels of risks for water contamination. However, the proximity to the source, the availability of sufficient water, and economics of the institution also play key roles.^35–37 A majority of zebrafish facilities rely on municipal water because they are located in urban areas. However, municipal water may also carry different types of pollutants, depending on the economic and geographic area. Thus, there exists intra- and interfacility variability for the quality of source water that needs to be controlled.^35–37

To this end, facilities eliminate pollutants in the municipal water by reverse osmosis water filtration systems (ROWS), which can significantly reduce, but not completely eliminate, the relative amount of total dissolved solids and pollutants. Freshwater conditions are reconstituted by adding commercially available synthetic sea salts, or laboratory-formulated salts and buffering agents such as sodium bicarbonate or calcium carbonate. While ROWS provide source water with reproducible quality, their main disadvantages are the high cost of maintenance and the amount of discharge water they generate.

RWFS are composed of a mechanical filter that removes coarse detritus and debris from food residue and feces that are generated in aquaria. A biological filter (containing sand, ceramic, or plastic beads as filter material) harbors nitrifying bacteria that oxidize ammonia and convert it into less toxic nitrites and nitrates. An aeration system or a trickle column provides gas exchange and establishes atmospheric levels of oxygen in water. Depending on the filter materials used in the mechanical and biological filters, fine sediment filters (FSFs) (containing zeolite or similar minerals) can be used to trap flocculants that would otherwise reduce the efficacy of UV water sterilizer units by shielding microorganisms from UV exposure. UV sterilizers are typically last in line before water flows back to the aquatic housing systems. These differ by manufacturers but are typically made of glass or polycarbonate. Pumps are required to move water between the filter components and the fish tanks. Various probes are placed in the water to monitor flow or quality, and water-heating systems are needed to establish species-specific water temperatures. Water recirculates constantly and a portion of it, typically 10%–20% every day, is exchanged with fresh, reconstituted source water.^36–40

With this study, we aim to determine the ranges of metal and metalloid concentrations for zebrafish in seven institutions: (1) the Champalimaud Center for the Unknown (CCU, Lisbon, Portugal), (2) the Centro Nacional de Investigaciones Cardiovasculares (CNIC, Madrid, Spain), (3) the European Zebrafish Resource Center (EZRC, KIT, Eggenstein-Leopoldshafen, Germany), (4) two CRWS in the Institute of Anatomy (IA) at the University of Bern (IA.1 and IA.2, Bern, Switzerland), (5) the Karolinska Institutet (KI, Stockholm, Sweden), (6) the Norwegian University of Life Sciences (NULS, Oslo, Norway), and (7) the Zebrafish International Resource Center (ZIRC, Eugene, USA). We analyzed 45 elements using inductively coupled plasma mass spectrometry (ICP-MS) and Ca using atomic absorbance spectrometry (AAS). Our evaluation of trace metals in different systems is unprecedented for zebrafish facilities and sheds light on some sources of elements in the water. Our findings help to establish a baseline and ranges for zebrafish water composition, provide strategies to troubleshoot problems with water quality, and consequently support reproducible animal welfare and husbandry conditions for zebrafish and other aquatic models.

Materials and Methods

Elemental analysis

Mercury analysis

Triplicates of fish water samples from the IA.1 and IA.2 were collected for mercury (Hg) analysis in 10 mL glass tubes. As diluents, concentrated acids were added for a final concentration of 1% nitric acid (HNO₃, 70%; #438073; Sigma) and 0.5% hydrochloric acid (HCl, 37%; #320331-500ML; Sigma). The samples were treated with HNO₃ and HCl as a diluent to keep the elements in suspension until they were analyzed in the ICP-MS. Then, Hg samples were filtered with 450 nm filters (#721-1345; Thermo Fisher). The instrument was calibrated with known Hg concentrations (0.05–2 μg L⁻¹; #28941-100ML-F; Sigma). Hg analysis was performed with a 7700x ICP-MS (Agilent Technologies, Santa Clara, USA). The ICP-MS was rinsed with three different chemical mixtures between each sample to prevent any carryover of Hg between samples.⁴¹

Ca analysis

Triplicates of fish water samples from three participating facilities (CNIC, IA.1, and ZIRC) were collected for Ca analysis in 50 mL Falcon tubes, concentrated HNO₃ (70%; #438073; Sigma) was added for a final concentration of 1% HNO₃ as diluent, and 50 μL of 2 g/L cesium chloride solution (#51869-250 mL; Sigma) was added as ionization buffer in all samples and standards. The instrument was calibrated with a standard solution of known Ca concentrations (0–10 mg L⁻¹; #1.19778.0500; Merck). Ca analysis was performed with a ZEEnit 700 P AAS (Analytic Jena AG, Jena, Germany). After every tenth sample, two of the standards (2.5 and 8 mg L⁻¹) were measured again to assess instrument stability and avoid drift.

Multielement analysis

Triplicates of municipal water, purified water, and fish water samples from all the participating facilities (CCU, CNIC, EZRC, IA.1, IA.2, KI, NULS, and ZIRC) were collected for multielement analysis of aluminum (Al), arsenic (As), barium (Ba), beryllium (Be), B, cadmium (Cd), cerium (Ce), cesium (Cs), chromium (Cr), Co, Cu, dysprosium (Dy), erbium (Er), europium (Eu), gadolinium (Gd), gallium (Ga), holmium (Ho), Fe, lanthanum (La), lead (Pb), lutetium (Lu), Mg, Mn, Mo, neodymium (Nd), nickel (Ni), K, praseodymium (Pr), rubidium (Rb), samarium (Sm), Se, silicon (Si), silver (Ag), Na, strontium (Sr), thallium (Tl), thorium (Th), thulium (Tm), titanium (Ti), uranium (U), ytterbium (Yb), vanadium (V), and Zn. K, Mg, Mo, Na, Si, and Ti were in 50 mL Falcon tubes. Concentrated nitric acid (HNO₃, 70%; #438073; Sigma) was added for a final concentration of 1% HNO₃. The samples were treated with HNO₃ as a diluent to keep the elements in suspension until they were analyzed in the ICP-MS. The instrument was calibrated with a multielement standard solution (500, 100, 10, 1, 0.1, and 0 μg L⁻¹; # IV-ICPMS-71A; Agilent).

Multielement analysis was performed with 7700x ICP-MS (Agilent Technologies). After every 10th sample, two of the standards (1 and 100 μg L⁻¹) were measured again to assess instrument stability and avoid drift. ¹⁰³Rh and ¹¹⁵In were used as internal standards during the measurements. Metal concentrations in each sample were determined according to appropriate standard curves obtained from the calibration of the corresponding metal standards.^42–44 The method is adapted from the EPA Method 3052.

Sample collection

Water samples were taken from different strategic access points of the water systems in this study. The water system for the fish facility at the IA.1 is shown as an example in Figure 1. These points of access included the following: municipal or source water (Fig. 1 (1), pre-ROWS), purified source water (Fig. 1 (2), post-ROWS), water from the bacteriological filter and/or mechanical filters (Fig. 1 (3)), and fish water (Fig. 1 (4)).

FIG. 1.

Schematic of a centralized, recirculating water system. Shown is a scheme of the CRWS, at the IA from Zebcare (IA.1 according to the nomenclature used in this article). Blue arrows represent direction of water flow from the municipal water source (1), through a prefilter unit and ROWS, to the reservoir or water tank (2). Yellow arrows indicate the recirculating, filtered system water between RWFS (3) and fish housing system (4). The red arrow indicates daily water removal to drain (15% of the total facility water volume). CRWS, centralized recirculating water systems; IA, Institute of Anatomy; ROWS, reverse osmosis water filtration systems; RWFS, recirculating water filtration system. Color images are available online.

All water samples from ZIRC, CCU, NULS, CNIC, KI, and EZRC were delivered to the IA after nitric acid treatment and stored at 4°C until analysis.

Water samples to determine the impact of ROWS filter membrane changes

Triplicates of municipal water and ROWS water samples were collected and analyzed before and 24 h after the change of 2-year-old membranes (#CDRC60202; Merck) at the IA.1. The specific model analyzed was the RiOs Essential 16 water purification system (#ZR0E0160WW; Merck).

Water samples for a temporal quality assessment

Over a 3-month period, fish water samples were collected weekly from fish tanks at the CNIC facility.

Water samples from biological filters

Triplicate samples from the various biological filters were collected at the following fish facilities: 1.

Propeller wash bead filter (PWBF) samples: ZIRC

FSF samples: ZIRC.

Bead moving filter (BMF) samples: CCU, IA.1, KI, and NULS.

List of water systems and technical specifications

Main water quality parameters, compounds added, and some technical specifications of the different water systems that contributed to this study are listed in Table 2.

Table 2.

Water Quality Parameters, Added Compounds, and Technical Specifications from Eight Fish Facilities

Centralized recirculating water systems
Institution	CCU	CNIC	EZRC	IA.1	IA.2	KI	NULS	ZIRC
Water quality
Temperature	28–28.5	28–28.5	27–27.5	28–28.5	28–28.5	28–28.5	28–28.5	28–28.5
Conductivity (μS/cm)	1200–1250	450–500	220	600–700	500–600	750–800	400–500	450–550
pH	7.4–7.6	7.4–7.6	7–7.2	7.4–7.6	7.4–7.6	6.8–7.2	7.5–7.7	7.5–7.6
External compounds added in the CRWS
Main diet	Sparos Zebrafeed 400–600 μm	Gemma Micron 500	Caviar and TetraMin flakes	Gemma Micron 500	Gemma Micron 300	Sparos Zebrafeed 400–600 μm	SDS400	Master mix
Live food	A and R	A	A	A	A	R	A	A and R
Synthetic sea salt	Sal basic plus	Sera marin	—	Reef crystals, aquatic systems	Instant ocean, aquatic systems	Tetra marine	Instant ocean, aquatic systems	Instant ocean, aquatic systems
Others	SB	SB	—	SB	SB	SB	SB	Ar and Zeo
Technical specifications
Brand CRWS	Tecniplast	Aquaneering	Aquaschwartz	Zebcare	Tecniplast	Tecniplast	Aquatic habitats	Aquaneering
Year of construction	2011	2010	2003	2015	2010	2018	2008	1999
Percentage of water change per day	10–15	10–15	10–15	15–20	10–15	10–15	10–15	10–15
Water source	MW	MW	GW (1/3)+MW (2/3)	MW	MW	MW	MW	MW
Water treatment; brand; model	ROWS; Merck Millipore; RiOS200	ROWS; Cuno; HP-WM-900	—	ROWS; Merck Millipore; RiOS16	ROWS; Merck Millipore; RiOS16	ROWS; E2 GE Osmonics	ROWS; E2 GE Osmonics	ROWS; Cuno; HP-WM-900
Type of filtration of water	Solids removal, chemical and biological	Solids removal, chemical and biological	Solids removal and biological	Mechanical and biological	Solids removal, chemical and biological	Solids removal, chemical and biological	Solids removal and biological	Solids removal, chemical and biological

Institutions: CCU, Champalimaud Center for the Unknown (Lisbon, Portugal); CNIC, Centro Nacional de Investigaciones Cardiovasculares (Madrid, Spain); EZRC, European Zebrafish Resource Center (Eggenstein-Leopoldshafen, Germany); IA.1 and IA.2, Institute of Anatomy (Bern, Switzerland); KI, Karolinska Institutet (Stockholm, Sweden); NULS, Norwegian University of Life Sciences (Oslo, Norway); ZIRC, Zebrafish International Resource Center (Eugene, USA). Main diet: Master mix (Zeigler, O.S.I and Golden Pearl). Live food: A, artemia; R, rotifer. Others: Ar, aragonite; SB, sodium bicarbonate; Zeo, zeolite. Water source and water treatment: GW, groundwater; MW, municipal water; ROWS, reverse osmotic water filtration system. Water filtration: solids removal (filter pads, fine sediment filter, drumfilter, or cartridge filters); chemical (activated carbon); biological (bead moving filter or propeller wash bead filter). CRWS, centralized recirculating water system.

Fish facilities

The participating facilities were established at least 1 year before our study. The following three criteria characterized the facilities that participated in sampling and water testing. (1) Health: During the duration of our study, specifically at times of sample collections, no disease outbreaks were reported in any of the facilities. The majority of the facilities carry out regular health status monitoring including microbial status. (2) Animal Environment: All facilities were operated under environmental conditions that fell into the range of recently published guidelines for zebrafish husbandry by Aleström et al.,⁴⁵ for FELASA (the European equivalent of AALAS). These conditions included temperature, pH, and conductivity and are listed in Table 2. (3) Facility Performance: In Table 3, we list the biological parameters that can be considered normal for most facilities for the purposes of this study; these include breeding success, fertilization rates, hatching rates, postnursery survival rates, as well as total population and stocking densities.

Table 3.

Biological Parameters from Fish Facilities Included in This Study

Biological parameters
Institutions	CCU	CNIC	EZRC	IA.1	IA.2	KI	ZIRC
Total fish populations	21,500	12,000	350,000	12,000	4000	15,000	52,500
Total no. of single crosses	30	21	19	20	20	14	135
Breeding success (%)	73	70	70	85	75	90	77% ± 15%
Total no. of eggs	3212	2786	671	1050	850	770	—
Fertilization rate (%)	94	86	88	90	90	80	76% ± 19%
Hatching rate: 48–72 hpf (%)	100	100	100	100	100	100	96% ± 2%
Survival rate at the time of entering the CRWS (%)	88	71	—	75	70	90	92.8% ± 7%
Stocking densities (fish/per liter)	10	5	5	5–6	5–6	5	4–7

The table shows different parameters to evaluate husbandry success at zebrafish facilities included in this study. Institutions: CCU; CNIC; EZRC; IA.1 and IA.2; KI; and ZIRC. Breeding success: number of pairwise crosses that produced eggs divided by the number of all crosses set up. Fertilization rate: number of fertilized eggs divided by the number of viable eggs laid. Fertilization is determined by cell divisions, and typically 0–4 h after fertilization. Hatching rate 48–72 hpf: number of hatched embryos (around 72 hpf) divided by the number of fertilized eggs. Survival rate after 30 dpf: number of larvae that survived the nursery period versus the number of fertilized embryos that were entered into the nursery. Stocking density: number of fish per liter.

Data management and statistics

GraphPad Prism version 7.04 was used for statistical analysis and charting of results (GraphPad Software, Inc., San Diego, CA). Statistical significance was determined using the Holm–Sidak method for multiple t-tests and Sidak's multiple comparisons test for two-way analysis of variance, with alpha = 0.05. Raw data have been deposited at Mendeley (10.17632/htdhxbpgpg.3).

Ethical statement

Animals were housed in accordance with bioethical regulations for the use of laboratory animals from the corresponding countries: CCU, Portuguese General Directorate of Veterinary (DGAV); CNIC, the Community of Madrid “Dirección General de Medio Ambiente,” Spain; ERZC, the Government of Baden-Württemberg, Regierungspräsidium Karlsruhe, Germany; IA.1 and IA.2, Amt für Veterinärwesen, Canton of Bern, Switzerland; KI, the international and local ethical guidelines; NULS, the Norwegian Food Inspection Authority (NFIA); ZIRC, the University of Oregon Institutional Animal Care and Use Committee.

Results

Metal and metalloid concentrations in eight recirculating water systems

To determine metal and metalloid concentrations in CRWS, we analyzed fish water from eight facilities distributed over seven countries (Germany, Norway, Portugal, Spain, Sweden, Switzerland, and the United States). The highest concentrations were detected for the macronutrients Ca, K, Na, and Mg (Table 4). Some nonessential elements, such as Al, B, Ba, Si, and Sr, were detected at higher concentrations than the essential elements, Fe and Zn. The elements As, Cr, Co, Cu, Mn, Mo, Ni, Pb, Rb, Ti, and V were detected in a range between 0.1 and 2 μg L⁻¹; whereas traces of Be, Cd, Cs, Ga, Hg, Pb, Se, Th, Tl, and U were found below 0.1 μg L⁻¹ in most fish facilities (Table 4). Some metal traces—Dy, Er, Eu, Gd, Ho, La, Lu, Nd, Pr, Sm, Tm, and Yb—were detected at very low concentrations or remained undetectable (Table 5).

Table 4.

Multielement Data Analysis from Eight Centralized Recirculating Water Systems

Institutions	CCU		CNIC		EZRC		IA.1		IA.2		KI		NULS		ZIRC
Elements	Mean	RSD	Mean	RSD	Mean	RSD	Mean	RSD	Mean	RSD	Mean	RSD	Mean	RSD	Mean	RSD
Macronutrients (mg L⁻¹)
Calcium	—	—	16	2.75	—	—	6.11	0.45	—	—	—	—	—	—	16.3	2.02
Magnesium	20.4	0.95	6.51	7.98	—	—	10.9	1.72	—	—	7.94	4.12	5.34	3.92	11	3.07
Potassium	7.57	0.28	4.87	7.34	—	—	7.07	26.22	—	—	3.50	6.79	2.85	5.22	4.3	2.22
Sodium	210	0.25	83.6	6.30	—	—	144	4.66	—	—	110	4.21	56.3	4.50	83.2	2.59
Essential elements (μg L⁻¹)
Chromium	0.18	31.04	0.34		0.03	12.36	0.15	4.11	0.26	36.26	0.06	37.69	0.05	9.96	0.04	15.33
Cobalt	0.09	—	0.02	—	ND	—	0.37	4.83	0.15	44.18	0.01	—	0.70	2.56	ND	—
Copper	0.59	57.60	ND	—	ND	—	3.16	2.59	0.24	40.86	1.54	127.5	0.14	59.95	ND	—
Iron	4.23	43.53	4.21	43.42	1.77	49.32	1.94	7.11	1.49	50.43	1.15	102.3	0.68	22.48	0.95	26.24
Manganese	0.52	87.15	0.17	91.41	0.13	25.87	0.11	11.24	0.19	50.79	ND	—	0.11	4.24	0.05	18.80
Molybdenum	0.70	2.76	0.25	14.54	—	—	0.34	2.95	—	—	—	—	—	—	0.28	12.19
Selenium	ND	—	0.39	18.34	—	—	ND	—	ND	—	ND	—	ND	—	0.52
Zinc	14.5	36.20	4.31	55.12	13.3	65.86	14.4	29.18	7.11	26.06	ND	—	4.51	6.43	ND	—
Nonessential elements (μg L⁻¹)
Aluminum (pH >6.5)	34.1	6.30	3.67	39.56	25.2	1.52	13.3	2.93	3.51	64.34	ND	—	ND	—	36.9	1.55
Arsenic	0.99	6.03	1.77	24.63	0.38	1.83	0.82	0.12	0.50	33.08	0.23	18.03	0.35	0.67	0.11	21.21
Barium	7.79	1.50	3.78	11.14	45.1	0.22	13.5	0.41	3.37	8.46	0.22	9.50	42.9	0.58	2.67	0.87
Beryllium	ND	—	0.01	—	0.01	—	0.01	—	0.24	286.43	0.01	—	0.01	—	0.04	—
Boron	79.1	3.12	7.38	8.43	5.49	7.40	55.4	4.23	35.09	43.48	12.6	8.16	14.9	3.88	35.6	3.98
Cadmium	0.01	—	0.01	—	ND	—	0.01	—	0.21	43.18	ND	—	0.01	—	ND	—
Cesium	0.15	1.60	0.02	—	0.01	—	0.01	3.43	0.12	35.47	0.01	—	0.01	—	0.05	1.47
Gallium	0.41	2.56	0.01	—	0.01	—	0	—	0.13	77.75	0	—	0.01	—	0.03	4.33
Lead	0.01	—	ND	—	ND	—	0.02	6.71	0.19	34.91	ND	—	0.01	—	ND	—
Mercury	—	—	—	—	—	—	0.02	2.72	0.02	5.44	—	—	—	—	—	—
Nickel	0.25	57	0.16	123.5	0.26	13.07	0.21	1.74	0.38	41.26	ND	—	0.36	0.64	ND	—
Rubidium	1.80	1.14	1.94	15.84	0.69	2.30	1.08	0.41	0.92	25.98	0.27	9.94	0.37	1.18	1.82	0.68
Silicon	1752	0.28	9358	6.16	—	—	—	—	440	4.05	360	13.22	546	0.24	2967	53.2
Silver	0.04	—	ND	—	0.18	102.90	ND	—	0.10	82.30	ND	—	ND	—	0.03	52.19
Strontium	206	0.19	38.2	10.54	132	0.24	37.9	1.58	70.26	18	43.9	3.35	19	2.24	154	0.35
Thallium	0.01	8.65	ND	—	ND	—	ND	—	0.07	67.66	ND	—	ND	—	ND	—
Thorium	0.07	—	0.10	116.87	0.02	5.61	0.03	13.92	0.14	46.69	0.05	—	0.03	—	ND	—
Titanium	0.44	1.22	0.37	25.29	—	—	—	—	0.28	2.76	0.36	15.68	0.42	12.80	0.33	23.64
Uranium	0.03	—	0.01	—	0.44	0.56	0.05	0.70	0.13	41.34	0.01	—	0	—	ND	—
Vanadium	1.40	0.98	0.80	24.71	0.11	6.78	0.41	3.28	0.31	30.16	0.19	11.04	0.13	1.66	0.37	1.84

Mean and RSD of metals and metalloid concentrations from eight CRWS: CCU; CNIC; EZRC; IA.1 and IA.2; KI; NULS; and ZIRC.

ND, not detected; RSD, relative standard deviation.

Table 5.

Trace Metals in Eight Centralized Recirculating Water Systems

Nonessential elements (μg L⁻¹)
Institutions	CCU	CNIC	EZRC	IA.1	IA.2	KI	NULS	ZIRC
Elements	Mean	Mean	Mean	Mean	Mean	Mean	Mean	Mean
Cerium	0.006	0.002	ND	ND	ND	ND	ND	ND
Dysprosium	ND	0.001	ND	ND	ND	ND	ND	ND
Erbium	0.001	ND	ND	ND	ND	ND	ND	ND
Europium	0.002	0.001	0.01	0.002	0.002	ND	0.007	ND
Gadolinium	0.002	ND	ND	ND	ND	ND	ND	ND
Holmium	ND	ND	ND	ND	ND	ND	ND	ND
Lanthanum	0.004	ND	ND	ND	ND	ND	ND	ND
Lutetium	ND	ND	ND	ND	ND	ND	ND	ND
Neodymium	0.003	0.002	ND	ND	ND	ND	ND	ND
Praseodymium	ND	ND	ND	ND	ND	ND	ND	ND
Promethium	ND	ND	ND	ND	ND	ND	ND	ND
Samarium	ND	ND	ND	ND	0.001	ND	ND	ND
Thulium	ND	ND	ND	ND	ND	ND	ND	ND
Ytterbium	0.002	ND	ND	ND	0.002	ND	ND	ND

Mean values of nonessential trace metal concentrations from eight CRWS. Institutions: CCU; CNIC; EZRC; IA.1 and IA.2; KI; NULS; ZIRC.

Our results indicate that the concentration of several nonessential elements (Al, B, Ba, Si, and Sr) was one or several orders of magnitude higher than that of essential elements such as Fe and Zn. Furthermore, we observed below technical threshold of detection, absence, or considerably low concentrations of several essential elements such as Cr, Co, Cu, Mn, Mo, and Se (Table 4).

Reverse osmotic water systems do not remove all elements with equal efficiency from municipal water

Municipal water from seven cities and their corresponding reverse osmosis purified water samples were analyzed for 38 elements to determine the capacity of ionic rejection in six fish facilities (CCU, CNIC, IA, KI, NULS, and ZIRC). Seven elements were detected in significantly higher concentrations than the other elements we analyzed: Al, B (except at NULS), Ba, Cu, Fe, Sr, and Zn (except at KI and NULS) (Fig. 2). The rest of the elements were detected at low concentrations (< 0.1 μg L⁻¹) or not traced at all.

FIG. 2.

Metal concentrations before and after reverse osmosis purification of water in six different fish facilities. Municipal water (n = 3) and reverse osmosis purified water (n = 3) samples were collected at CCU, CNIC, IA.1, KI, NULS, and ZIRC. Data are presented as mean ± SD; *p < 0.05; and ***p < 0.001 according to multiple t-tests. CCU, Champalimaud Center for the Unknown; CNIC, Centro Nacional de Investigaciones Cardiovasculares; KI, Karolinska Institutet; NULS, Norwegian University of Life Sciences; ND, not detected; ZIRC, Zebrafish International Resource Center.

To test whether effective ion rejection is associated with the efficiency of reverse osmosis filtration or related to the regular maintenance of the exchange of membranes, we analyzed municipal and purified water samples before and after the change of membranes in one facility. After replacement, we detected efficient ionic rejection only for the removal of B (Fig. 3). For other elements, the membrane change did not appear to make any significant improvement for product water quality.

FIG. 3.

Percentage of ionic rejection before and after the change of reverse osmosis membranes. Municipal water (n = 3) and reverse osmosis purified water (n = 3) samples were collected at IA.1 before and 24 h after the change of 2-year-old membranes. The ROWS is maintained at regular intervals. Data are presented as mean ± SD; **p < 0.01 according to multiple t-tests. AMC, after membrane change; BMC, before membrane change.

Our results indicate a considerable presence of Al, B, and Fe after the filtration of municipal water through ROWS and that the replacement of 2-year-old membranes reduced mainly B. Thus, ionic rejection across ROWS is a differential process.

Metal concentrations fluctuate over a period of 3 months

Fish water samples were collected over a 3-month period to evaluate the fluctuation of metals in the CNIC fish facility. While for most elements concentrations remained stable; Al, As, Fe, V, and Zn showed fluctuations higher than 25% (Fig. 4). In contrast, the concentrations of V and As remained consistently low and hardly fluctuated at all.

FIG. 4.

Fluctuation of metal concentrations over a period of 3 months. Fish water samples (n = 14) were collected at CNIC. Graph shows the fluctuation of Al, As, Fe, V, and Zn levels (μg L⁻¹) over a period of 3 months in the zebrafish housing system. Color images are available online.

The comparison between purified water and fish water concentrations (Fig. 2 and Table 4) with levels of fluctuation over the 3-month tracking period (Fig. 4) suggests that most of the variability of these elements reflects addition of factors that varied during a 3-month tracking. For example, we detected Zn at 11.28 μg L⁻¹ in the municipal water and 0.38 μg L⁻¹ after ROWS filtration at CNIC (Fig. 2). On average, we detected a Zn concentration with a minimal value of 1.54 and a maximum value of 8.24 μg L⁻¹ in the fish water samples over the 3-month period. This suggests that the addition of external factors to the water system, such as diet or synthetic salts, which contain Zn and Fe but presumably no or little As or V, contributes to the variability of some metal and metalloid concentrations in fish water.

Fine sediment and biological filters can accumulate metals

To test the performance of biological filters and FSF with the capacity to accumulate metals and metalloids, we compared fish water with the sump water from a PWBF (ZIRC), an FSF (ZIRC) and a BMF (CCU, IA.1, KI, and NULS) (Fig. 5). Water from the PWBF showed a significant accumulation of Al, Ba, Cu, Fe, Mn, Sr, and Zn concentrations when compared with fish water (Fig. 5A). Similarly, FSF water also showed a significant increase of Cu, Fe, Mn, and Sr concentrations when compared with fish water (Fig. 5A). In contrast, we did not detect any significant difference in metal concentrations between pre- and post-BMF water samples, except for Al and Mn at the KI facility (Fig. 5B).

FIG. 5.

Metal concentrations are elevated in PWBF and FSF samples relative to system water, but not in BMF samples. (A) Water samples from fish tanks (n = 3), PWBF (n = 3), and FSF (n = 3) were collected at ZIRC. (B) Water samples from fish tanks (n = 3) and BMF (n = 3) were collected at CCU, IA.1, KI, and NULS. Data are presented as mean ± SD (n = 3); *p < 0.05; and ***p < 0.001 according to multiple t-tests (A) and two-way ANOVA (B). ANOVA, analysis of variance; BMF, bead moving filter; FSF, fine sediment filter; PWBF, propeller wash bead filter.

Our results indicate a significant accumulation of Al, Ba, Cu, Fe, Mn, Sr, and Zn in PWBF and FSF in the ZIRC water system; while BMF does not show significant accumulation of any element in four fish facilities (CCU, IA.1, KI, and NULS).

Discussion

Controlled environmental conditions for laboratory animals provide a basis for successful animal husbandry and welfare and reproducible results. For zebrafish, husbandry studies included the roles of anesthesia and euthanasia,⁴⁶ diseases and pathogens,^47,48 behavior,^{20–23,49,50} housing,^{39,45,49–51} nutrition,^52,53 and reproduction.^54–56

In this study, we explored concentrations of metals and metalloids in different CRWS that are used in biomedical zebrafish research facilities. In this regard, our study could serve as a reference for troubleshooting when there are indications of diseases related to deficiency or excess of some elements.

Analysis of macro- and micronutrients and nonessential elements

We detected the macronutrients Ca, K, Mg, and Na at high concentrations, but barely detected the micronutrients Cr, Co, Cu, Fe, Mn, Se, and Zn. The Cu, Mn, and Zn metal content in early embryos has been suggested to be set by the maternal contribution and only increases once zebrafish develop to the stage where they can acquire nutrients through the diet or the environment.⁵⁷ The concentrations of Cu, Mn, and Zn reported in the study of Thomason et al. indicated that relatively low concentrations are required for fish younger than 30 dpf, suggesting that the concentrations of the CRWS shown in this study might be sufficient for normal biological processes in larval and juvenile zebrafish.

In several CRWS, we also detected elements that are described in the literature as nonessential (Al, B, Si, and Sr) and even toxic at excessive levels in concentrations higher than 1 μg L⁻¹. Al, for example, has been reported to increase the acetylcholinesterase activity in the zebrafish brain at 50 μg L⁻¹ AlCl₃ at pH 5.8²¹ and to induce behavioral changes at 6700–26,700 μg L⁻¹ AlCl₃.⁵⁸ In addition, neurotoxic effects have been reported in astroglia at 13,300 μg L⁻¹ AlCl₃.⁵⁹ Sr, as a further example, stimulates the process of bone mineralization at 520 μg L⁻¹ SrCl₂, and the addition of strontium citrate as a nutritional supplement can increase bone mineral density.⁵³ However, at concentrations exceeding 52,300 μg L⁻¹, SrCl₂ can also lead to the inhibition of the mineralization process during embryonic osteogenesis.⁶⁰ For the two other elements, B⁶¹ and Si, little is known, for example, about the tolerance levels, or acute and chronic toxicity in zebrafish. Thus, depending on the concentrations, nonessential elements can have beneficial or detrimental effects on zebrafish health.

Despite RO, RWFSs form dynamic animal environments

Evidence for differences of chemical composition of fish water can be gained from two facilities at the same institution (IA.1 and IA.2; see Table 4) that share the same ROWS source of water, but the overall infrastructure, synthetic sea salts, and diets differ between the two facilities. These data can be valuable and taken into account when conducting comparative studies between institutions that work in the same field of research.

Few studies have shown whether or not fluctuating environmental conditions such as temperature, conductivity, pH, hardness, and oxygen, parameters that play a role for the speciation of metal ions, influence metal and metalloid toxicities.^2,4,27,62 Therefore, additional information obtained by careful control of environmental parameters for each of these elements is needed to better understand under which conditions environmental changes and metal concentrations might cause chronic, acute, subpathological, or apparent disorders of behavior, metabolism, reproduction, health, and development in zebrafish.^{16,19–22,26,27,31,33,58,59,63–66}

Seven of the analyzed CRWS use municipal water as their water source that is purified by ROWS and reconstituted with synthetic sea salts and buffers to reach freshwater conditions. The efficiency of ROWS to filter water appears to be similar between facilities based on the relative concentrations of dissolved elements present in their municipal water. However, our observations suggest that some elements are removed more efficiently (Ba and Sr) than others (mainly Al, B, and Fe) by ROWS filter membranes. The causes of the observed differences in ionic rejection between elements may be due to excessive convective transport, poor size exclusion, or ineffective charge repulsion.⁶⁷ The concentration of particular elements after ROWS filtration can be used as an indication that membranes need to be replaced periodically. Previous studies already reported about the efficiency of ROWS for removing heavy metals,^68,69 but there is lack of and occasionally confounding information regarding the most adequate timing to replace ROWS filtration membranes. Therefore, we suggest more long-term studies to better understand optimal, facility-specific frequencies for filter and membrane changes and the impact of seasonal fluctuations on CRWS. Such measurements will contribute to more stable water composition, and thus, a reduction of costs for water system consumables, environmental sustainability, and the increased knowledge about the chemical composition of CRWS water. Despite ROWS and reconstituted freshwater conditions, we observed fluctuations of Al, Fe, and Zn concentrations in one of the water systems over a 3-month period. This suggests that even with relatively tight control of source water quality, fish diets, fluctuation of the number of fish on the system, and buffering agents may introduce significant variations in system water quality.^70–72

Possible role of filtration components for system water composition and stability

Biological filters represent crucial infrastructural components of RWFS for the oxidation of ammonia into nitrite and subsequently to less toxic nitrate.^38,39 Moreover, bacteria, algae, and fungi have the capacity to chelate metals.^73,74 However, so far there are no reports indicating which elements are accumulated by microorganisms in fish facilities. In addition, biofilm appears to be more effective in trapping metals in PWBF in comparison with BMF. We suggest that the more abrasive backwashing in BMF knocks off and removes microorganism-containing biofilm that is beneficial for the elimination of metals from immediate contact with fish. Moreover, it has not been reported yet, whether or not filter systems may also function as metal traps in RWFS. Specific studies are required to better understand which microorganisms might be involved in this process, which specific metals they accumulate, and how they impact the overall water quality in CRWS.⁷⁵ Several factors might affect metal accumulation in filter components: the population growth phase of the bacterial population, the type of materials used in the filters (beads, minerals, pads, etc.), the total biomass of the fish population relative to the total water volume of the system, and the level of its pollutant equilibrium.

In conclusion, our study sheds light on 46 elements that can be expected in CRWS. This information might be helpful for the research community as a troubleshooting aid to solve potential husbandry problems related to changes in water quality.

Footnotes

Acknowledgments

We thank the Mercader lab, Valentin Djonov, Helmut Segner, Heike Schmidt-Posthaus, Almut Köhler, Uwe Strähle, Audrius Gegeckas, Aleström lab. We are also grateful to Nuria Basdedios, Andy Oates, Helmut Signer, Stephan Fisher, Daniele Soroldoni, Luis Cesar Fernandez, Donatello Serio, and Juan Ramos. We thank the members of the animal facilities of ZIRC, EZRC, Universität Bern, CNIC, Norwegian University of Life Sciences, Karolinska Institute, and Champalimaud Foundation.

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was partially funded by the European Research Council ERC Consolidator grant 819717, Congento Lisboa-01-0145-FEDER-022170, cofinanced by FCT (Portugal) and Lisboa2020, under the Portugal2020 agreement (European Regional Development Fund). D.L. and Z.M.V. of the ZIRC gratefully acknowledge support by the NIH/ORIP (P40 OD011021).

References

Enghag

: Encyclopedia of the Elements: Technical

Data-History-Processing-Applications

. John Wiley & Sons, Federal Republic of Germany, 2008.

Boyd

: Water Quality: An Introduction Springer Nature Switzerland AG, Cham, Switzerland, 2019.

Roberts

: Fish

Pathology

. John Wiley & Sons, Chichester, United Kingdom, 2012.

Wood Chris

, Farrell Anthony

, Brauner Colin

: Fish Physiology: Homeostasis and Toxicology of Essential Metals, Vol. 31. Academic Press, London, 2011.

Maret

. The metals in the biological periodic system of the elements: concepts and conjectures. Int J Mol Sci, 2016; 17:66.

Niyogi

, Nadella

, Wood

. Interactive effects of waterborne metals in binary mixtures on short-term gill-metal binding and ion uptake in rainbow trout (Oncorhynchus mykiss). Aquat Toxicol, 2015; 165:109–119.

Bury

, Walker

, Glover

. Nutritive metal uptake in teleost fish. J Exp Biol, 2003; 206:11–23.

Horng

, Yu

, Liu

, Chen

, Lin

. Potassium regulation in medaka (Oryzias latipes) larvae acclimated to fresh water: passive uptake and active secretion by the skin cells. Sci Rep, 2017; 7:16215.

McCormick

, Hasegawa

, Hirano

. Calcium uptake in the skin of a freshwater teleost. Proc Natl Acad Sci U S A, 1992; 89:3635–3638.

10.

Lin

, Hwang

. The control of calcium metabolism in Zebrafish (Danio rerio). Int J Mol Sci, 2016; 17.

11.

Martinez-Finley

, Chakraborty

, Fretham

, Aschner

. Cellular transport and homeostasis of essential and nonessential metals. Metallomics, 2012; 4:593–605.

12.

Olsson

P-E

, Kling

, Hogstrand

. Mechanisms of heavy metal accumulation and toxicity in fish. In: Metal Metabolism in Aquatic Environments. Depledge MH and Sanders B (eds), pp. 321–350. Chapman & Hall, London, United Kingdom, 1998.

13.

Beaver

, Nkrumah-Elie

, Truong

, et al. Adverse effects of parental zinc deficiency on metal homeostasis and embryonic development in a zebrafish model. J Nutr Biochem, 2017; 43:78–87.

14.

, Liu

, Zhong

, et al. Iron deficiency anemia's effect on bone formation in zebrafish mutant. Biochem Biophys Res Commun, 2016; 475:271–276.

15.

Salvaggio

, Marino

, Albano

, et al. Toxic effects of zinc chloride on the bone development in Danio rerio (Hamilton, 1822). Front Physiol, 2016; 7:153.

16.

Sfakianakis

, Renieri

, Kentouri

, Tsatsakis

. Effect of heavy metals on fish larvae deformities: a review. Environ Res, 2015; 137:246–255.

17.

, Zhang

, et al. Copper impairs zebrafish swimbladder development by down-regulating Wnt signaling. Aquat Toxicol, 2017; 192:155–164.

18.

Altenhofen

, Nabinger

, Pereira

TCB

, Leite

, Bogo

, Bonan

. Manganese(II) chloride alters nucleotide and nucleoside catabolism in zebrafish (Danio rerio) adult brain. Mol Neurobiol, 2018; 55:3866–3874.

19.

Augustine

, Gagnaire

, Adam-Guillermin

, Kooijman

. Effects of uranium on the metabolism of zebrafish, Danio rerio. Aquat Toxicol, 2012; 118–119:9–26.

20.

Zimmermann Prado Rodrigues

, Staudt

LBM

, et al. Histopathological, genotoxic, and behavioral damages induced by manganese (II) in adult zebrafish. Chemosphere, 2020; 244:125550.

21.

Senger

, Seibt

, Ghisleni

, Dias

, Bogo

, Bonan

. Aluminum exposure alters behavioral parameters and increases acetylcholinesterase activity in zebrafish (Danio rerio) brain. Cell Biol Toxicol, 2011; 27:199–205.

22.

Pilehvar

, Town

, Blust

. The interactive effect of copper(II) and conspecific alarm substances on behavioural responses of zebrafish (Danio rerio). Behav Brain Res, 2020; 381:112452.

23.

Toni

, Angiulli

, Miccoli

, et al. Environmental temperature variation affects brain protein expression and cognitive abilities in adult zebrafish (Danio rerio): a proteomic and behavioural study. J Proteomics, 2019; 204:103396.

24.

Christou

, Iliopoulou

, Witten

, Koumoundouros

. Segmentation pattern of zebrafish caudal fin is affected by developmental temperature and defined by multiple fusions between segments. J Exp Zool B Mol Dev Evol, 2018; 330:330–340.

25.

Cortemeglia

, Beitinger

. Temperature tolerances of wild-type and red transgenic zebra danios. Trans Am Fish Soc, 2005; 134:1431–1437.

26.

Delomas

, Dabrowski

. Larval rearing of zebrafish at suboptimal temperatures. J Therm Biol, 2018; 74:170–173.

27.

Dave

. The influence of pH on the toxicity of aluminum, cadmium, and iron to eggs and larvae of the zebrafish, Brachydanio rerio. Ecotoxicol Environ Saf, 1985; 10:253–267.

28.

Komjarova

, Blust

. Effects of Na, Ca, and pH on the simultaneous uptake of Cd, Cu, Ni, Pb, and Zn in the zebrafish Danio rerio: a stable isotope experiment. Environ Sci Technol, 2009; 43:7958–7963.

29.

Nugegoda

, Kibria

. Water quality guidelines for the protection of aquatic ecosystems. Encycl Aquat Ecotoxicol, 2013; 10:978–994.

30.

Acosta

DDS

, Danielle

, Altenhofen

, et al. Copper at low levels impairs memory of adult zebrafish (Danio rerio) and affects swimming performance of larvae. Comp Biochem Physiol C Toxicol Pharmacol, 2016; 185–186:122–130.

31.

Cui

, Zhang

, Bai

, et al. Effects of rare earth elements La and Yb on the morphological and functional development of zebrafish embryos. J Environ Sci (China), 2012; 24:209–213.

32.

Dahan

, Jude

, Lamendella

, Keesing

, Perron

. Exposure to arsenic alters the microbiome of larval zebrafish. Front Microbiol, 2018; 9:1323.

33.

Kwon

, Ha

, Jung

, et al. Effects of barium chloride exposure on hormones and genes of the hypothalamic-pituitary-gonad axis, and reproduction of zebrafish (Danio rerio). Bull Environ Contam Toxicol, 2016; 96:341–346.

34.

Xiao

, Cui

, Ding

, Wang

. Effects of lithium and 2,4-dichlorophenol on zebrafish: circadian rhythm disorder and molecular effects. Zebrafish, 2017; 14:209–215.

35.

Hammer

: Water quality for zebrafish culture. In: The Zebrafish in Biomedical Research. Cartner SC, Eisen JS, Farmer SC, Guillemin KJ, Kent ML and Sanders GE (eds), pp. 321–335. Academic Press, London, United Kingdom, 2020.

36.

Timmons

, Guerdat

, Vinci

. Recirculating Aquaculture. Ithaca Publishing Company LLC, New York, 2018.

37.

Trevarrow

. Zebrafish facilities for small and large laboratories. Methods Cell Biol, 2004; 77:565–591.

38.

Varga

. Aquaculture, husbandry, and shipping at the Zebrafish International Resource Center. Methods Cell Biol, 2016; 135:509–534.

39.

Lawrence

, Mason

. Zebrafish housing systems: a review of basic operating principles and considerations for design and functionality. ILAR J, 2012; 53:179–191.

40.

Courtland

. Recirculating systems for zebrafish. Lab Anim (NY), 2002; 31:53–56.

41.

Gygax

, Gfeller

, Wilcke

, Mestrot

. Emerging investigator series: mercury mobility and methylmercury formation in a contaminated agricultural flood plain: influence of flooding and manure addition. Environ Sci Process Impacts, 2019; 21:2008–2019.

42.

Fernández-Turiel

, Llorens

, López-Vera

, Gómez-Artola

, Morell

, Gimeno

. Strategy for water analysis using ICP-MS. Fresenius J Anal Chem, 2000; 368:601–606.

43.

Khosravi

, Zarei

, Bigalke

. Characterizing major controls on spatial and seasonal variations in chemical composition of surface and pore brine of Maharlu Lake, Southern Iran. Aquat Geochem, 2018; 24:27–54.

44.

Khosravi

, Zarei

, Sracek

, Bigalke

. Geochemical and hydrological controls of arsenic concentrations across the sediment–water interface at Maharlu Lake, Southern Iran. Appl Geochem, 2019; 102:88–101.

45.

Aleström

, D'Angelo

, Midtlyng

, et al. Zebrafish: housing and husbandry recommendations. Lab Anim, 2020; 54:213–224.

46.

Matthews

, Varga

. Anesthesia and euthanasia in zebrafish. ILAR J, 2012; 53:192–204.

47.

Kent

, Feist

, Harper

, et al. Recommendations for control of pathogens and infectious diseases in fish research facilities. Comp Biochem Physiol C Toxicol Pharmacol, 2009; 149:240–248.

48.

Sanders

, Watral

, Kent

. Microsporidiosis in zebrafish research facilities. ILAR J, 2012; 53:106–113.

49.

Collymore

, Tolwani

, Rasmussen

. The behavioral effects of single housing and environmental enrichment on adult zebrafish (Danio rerio). J Am Assoc Lab Anim Sci, 2015; 54:280–285.

50.

Parker

, Millington

, Combe

, Brennan

. Housing conditions differentially affect physiological and behavioural stress responses of zebrafish, as well as the response to anxiolytics. PLoS One, 2012; 7:e34992.

51.

Kurtzman

, Craig

, Grizzle

, Hove

. Sexually segregated housing results in improved early larval survival in zebrafish. Lab Anim (NY), 2010; 39:183–189.

52.

Watts

, Powell

, D'Abramo

. Fundamental approaches to the study of zebrafish nutrition. ILAR J, 2012; 53:144–160.

53.

Siccardi

, 3rd, Padgett-Vasquez

, Garris

, Nagy

, D'Abramo

, Watts

. Dietary strontium increases bone mineral density in intact zebrafish (Danio rerio): a potential model system for bone research. Zebrafish, 2010; 7:267–273.

54.

Fowler

, Williams

, Dennis-Cornelius

, et al. Influence of commercial and laboratory diets on growth, body composition, and reproduction in the zebrafish Danio rerio. Zebrafish, 2019; 16:508–521.

55.

Markovich

, Rizzuto

, Brown

. Diet affects spawning in zebrafish. Zebrafish, 2007; 4:69–74.

56.

Meinelt

, Schulz

, Wirth

. The fatty acid composition of diet influences the quality of fish eggs. J Appe Ichtyol, 1999; 15:13–19.

57.

Thomason

, Pettiglio

, Herrera

, Kao

, Gitlin

, Bartnikas

. Characterization of trace metal content in the developing zebrafish embryo. PLoS One, 2017; 12:e0179318.

58.

Capriello

, Grimaldi

, Cofone

, D'Aniello

, Ferrandino

. Effects of aluminium and cadmium on hatching and swimming ability in developing zebrafish. Chemosphere, 2019; 222:243–249.

59.

Monaco

, Grimaldi

, Ferrandino

. Aluminium chloride-induced toxicity in zebrafish larvae. J Fish Dis, 2017; 40:629–635.

60.

Pasqualetti

, Banfi

, Mariotti

. The effects of strontium on skeletal development in zebrafish embryo. J Trace Elem Med Biol, 2013; 27:375–379.

61.

Soucek

, Dickinson

, Koch

. Acute and chronic toxicity of boron to a variety of freshwater organisms. Environ Toxicol Chem, 2011; 30:1906–1914.

62.

Boyle

, Clark

, Handy

. Toxicities of copper oxide nanomaterial and copper sulphate in early life stage zebrafish: effects of pH and intermittent pulse exposure. Ecotoxicol Environ Saf, 2020; 190:109985.

63.

Jezierska

, Ługowska

, Witeska

. The effects of heavy metals on embryonic development of fish (a review). Fish Physiol Biochem, 2009; 35:625–640.

64.

Johnson

, Carew

, Sloman

. The effects of copper on the morphological and functional development of zebrafish embryos. Aquat Toxicol, 2007; 84:431–438.

65.

Tye

, Montgomery

, Hobbs

, Vanpelt

, Masino

. An adult zebrafish diet contaminated with chromium reduces the viability of progeny. Zebrafish, 2018; 15:179–187.

66.

Zheng

, Yuan

, Wu

, Lv

, Zhu

. Circadian time-dependent antioxidant and inflammatory responses to acute cadmium exposure in the brain of zebrafish. Aquat Toxicol, 2017; 182:113–119.

67.

Cui

, Ge

, Liu

X-Y

, Chung

T-S

. Novel forward osmosis process to effectively remove heavy metal ions. J Membr Sci, 2014; 467:188–194.

68.

Qdais

, Moussa

. Removal of heavy metals from wastewater by membrane processes: a comparative study. Desalination, 2004; 164:105–110.

69.

Thaçi

, Gashi

. Reverse osmosis removal of heavy metals from wastewater effluents using biowaste materials pretreatment. Pol J Environ Stud, 2019; 28:337–341.

70.

Sloth

, Julshamn

, Lundebye

. Total arsenic and inorganic arsenic content in Norwegian fish feed products. Aquacult Nutr, 2005; 11:61–66.

71.

Sele

, Sloth

, Holmelid

, Valdersnes

, Skov

, Amlund

. Arsenic-containing fatty acids and hydrocarbons in marine oils—determination using reversed-phase HPLC-ICP-MS and HPLC-qTOF-MS. Talanta, 2014; 121:89–96.

72.

Wang

W-X

. Dietary toxicity of metals in aquatic animals: recent studies and perspectives. Chin Sci Bull, 2013; 58:203–213.

73.

Gadd

. Heavy metal accumulation by bacteria and other microorganisms. Experientia, 1990; 46:834–840.

74.

Donati

. Heavy Metals in the Environment: Microorganisms and Bioremediation. CRC Press, Boca Raton, FL, 2018.

75.

Ayangbenro

, Babalola

. A new strategy for heavy metal polluted environments: a review of microbial biosorbents. Int J Environ Res Public Health, 2017; 14:94.

76.

https://www.waterquality.gov.au/sites/default/files/documents/anzecc-armcanz-2000-guidelines-vol1.pdf (accessed November 6, 2020).

77.

https://www.epa.gov/wqc/national-recommended-water-quality-criteria-aquatic-life-criteria-table#table (accessed November 6, 2020).